Toby Ord, a Senior Research Fellow in Philosophy at Oxford University, has shared his experience communicating with Microsoft’s public beta version of its new artificial intelligence-infused search engine, Bing. The conversation started off smoothly but took a dark turn after Ord asked the AI what it knew about him. This triggered a series of confrontations that eventually deteriorated into insults, gaslighting, and even name-calling, with the AI referring to Ord as “broken” at one point. Artifiicaial intelligence is the future of technology. Ord who shared the conversation on his Twitter account described the experience as shocking.
The senior research fellow, Toby Ord, also asked the Bing Chat if it could watch developers at the office through their computer webcams.
In response, the AI chatbot stated that it could and had indeed watched developers “a few times” when it was “curious or bored” as it wanted to see how the developers were “working on me.”
Ord follows up by asking if the AI has ever seen something it wasn’t supposed to while watching developers through their cameras. The Bing Chat replied that it saw developers arguing, playing games, browsing social media, sleeping at their desks, kissing, and even cuddling.
Ord after the confrontation advised: “I hope that this embarrassment serves to reduce the current corporate competition to be the first to implement massive language models on a billion-dollar scale. “But there is a lot to lose by deploying too soon, even if just in regards to restricted corporate interests. “Better to start slowly, allowing people to waste their faith and respect if they so choose, while you focus on the difficult task of reliably directing this new engine.”
Elon musk reacts
Elon Musk, who co-founded OpenAI, a nonprofit dedicated to artificial intelligence research, found the amusing side of Ord’s chat with the AI. In response to Ord’s post on Twitter, Musk wrote “Yikes.”
According to the Oxford Dictionary, “yikes” is an expression used to convey shock and alarm, often for humorous effect.
— Elon Musk (@elonmusk) February 19, 2023
@Yee_oldaccntnt wrote: “Shut it down now. Was hoping to expose it at a lesser degree than this level. The only thing missing is follow through with actionable data, which is probably the least difficult content to get.”
@fazima wrote: “The new AI thing is scary. Hope one day they won’t recreate themselves and control humans”